09HILKE FORMATTED
3/31/2006 9:25:39 AM
COMMENTARY
COMMENTS ON PETER CARSTENSEN’S
“CREATING WORKABLY COMPETITIVE
WHOLESALE MARKETS IN ENERGY”
John C. Hilke*
TABLE OF CONTENTS
I.
INTRODUCTION ................................................................... 156
II.
HISTORICAL IRONY ............................................................. 158
III.
COMMENTARY .................................................................... 160
A. Conditions Other than the Number of Market
Participants Are Critically Important in
Addressing Market Power Concerns .......................... 160
B. Other Ways of Increasing the Number of Market
Participants................................................................. 164
IV.
CONCLUSION ...................................................................... 165
V.
ADDENDUM ........................................................................ 166
A. Several Suppliers: Why Low Aggregate
Concentration Is Likely to Be Necessary If There
Are to Be Two or More Suppliers Competing to Set
the Market Clearing Price in Each Time Period........ 166
*
Ph.D. in economics and public policy, Cornell University. Dr. Hilke is an
independent consultant who previously served as an economist and the electricity project
director in the Federal Trade Commission’s Bureau of Economics. The views expressed
are the Author’s personal views and do not purport to be the views of the Federal Trade
Commission or of any individual Commissioner. The Author wishes to thank Timothy
Brennan and colleagues at the FTC, Michael Wroblewski, John Seesel, Denis Breen and
Luke Froeb.
155
09HILKE FORMATTED
156
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
[1:1
B. Vertical Separation: Supply from Distant
Generators Requires Transmission Access ................ 168
I.
INTRODUCTION
Professor Carstensen’s article says, in essence, that policy
makers have failed to structure U.S. electricity markets to
support competition, that current regulatory reforms are of
dubious value to consumers, and that the prospects for
improvement are not bright. My view is that this assessment is
too pessimistic because it defines structural improvements too
narrowly. Also, several readily identified and available reforms
can help address many concerns about existing market power in
electricity markets, and several states are already acting to
implement these reforms.1
In many ways, this comment is a different take on the same
questions that Diana Moss addressed in her highly useful
electricity market power literature review: what are the pressing
market power problems in electricity markets and how are we
2
doing, as a nation, in dealing with them? Professor Carstensen
aids his analysis of these questions with a parallel treatment of
market power concerns in the U.S. natural gas industry. This
parallel treatment is apt because many of the reformers from the
natural gas industry are active in the electric power industry
reforms,3 and there is some convergence between electricity and
natural gas markets.4
1.
These Comments focus on the electricity industry and are not necessarily
applicable to markets in other industries. Because this is a Comment rather than a
primary Article, it is assumed that readers are familiar with Professor Carstensen’s
Article, important peculiarities of electricity markets, and associated policy issues. Useful
references in this regard include: STEVEN STOFT, POWER SYSTEM ECONOMICS: DESIGNING
MARKETS FOR ELECTRICITY (IEEE Press and Wiley-Interscience 2002); TIMOTHY J.
BRENNAN, K AREN L. PALMER & SALVADOR A. MARTINEZ, A LTERNATING CURRENTS :
ELECTRICITY MARKETS AND PUBLIC POLICY (Resources for the Future 2002); S ALLY HUNT,
MAKING COMPETITION WORK IN ELECTRICITY (John Wiley & Sons, Inc. 2002); FED . TRADE
COMM’N, COMPETITION AND CONSUMER PROTECTION PERSPECTIVES ON ELECTRIC POWER
REGULATORY R EFORM: FOCUS ON R ETAIL COMPETITION (Sep. 2001); F ED. TRADE COMM’N,
COMPETITION AND CONSUMER PROTECTION PERSPECTIVES ON ELECTRIC POWER
REGULATORY REFORM (Jul. 2000).
2.
Diana Moss, Electricity and Market Power: Current Issues for Restructuring
Markets (A Survey), 1 ENVTL. & ENERGY L. & POL’Y J. 11 (2006).
3.
For example, the Gas Industry Standards Board decided to broaden its scope to
include electric power markets as well as natural gas markets. Accordingly, it has
changed its name to the North American Energy Standards Board. See Press Release,
North American Standards Board, Bylaws for North American Standards Board (Dec. 6,
2001), available at http://www.naesb.org/pdf/120601pr.pdf.
4.
For example, the settlement of the FTC’s challenge to a merger between DTE
and MichCon, respectively the electricity and natural gas distribution firms in the Detroit
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
157
One way to read Professor Carstensen’s article is as an effort
to return to the basic structure-conduct-performance paradigm of
industrial organization economics in examining market power in
electricity markets and in evaluating U.S. regulatory reform
efforts to date. The basic propositions that Professor Carstensen
emphasizes are (1) that electricity markets should have as many
suppliers as technically feasible, and (2) if there are stages of
production where a competitive structure is unlikely, vertical
integration between the monopolized stages and potentially
competitive stages of production is problematic for existing
competition and for effective entry at the potentially competitive
stages.5 For electricity markets, I generally agree that moving
away from a structure with a single generating firm in an area is
likely to promote competition, but, in my view, Professor
Carstensen goes too far by suggesting that an atomistic market
structure is necessary for effective competition in these markets.6
I generally agree that transmission access (unbundling) is
necessary for distant generators to compete with generators
located close to a load because transmission overbuilds are highly
unlikely and there is no alternative electric power delivery
technology on the horizon to the best of my knowledge.7 For this
reason, transmission is fully regulated and is likely to remain so.
With regard to vertical unbundling more generally, I think that a
real contribution of Professor Carstensen’s paper is bringing
attention to the historical unbundling decisions in rail
transportation that resulted in prohibitions against railroads
owning the commodities that they carry.8
area, focused on the loss of actual and potential competition between electricity and
natural gas to serve customers contemplating on-site generation of electric power fueled
by natural gas. See John C. Hilke, Convergence Mergers: A New Competitive Settlement
Model from Detroit, ELEC. J., Oct. 2001, at 13–18.
5.
Peter Carstensen, Creating Workably Competitive Wholesale Markets in Energy:
Necessary Conditions, Structure, and Conduct, 1 ENVTL. & ENERGY L. & POL’Y J. 85, 119–
20 (2006).
6.
The addendum explains why several generators in an area are likely to be
necessary to have effective competition in different time periods, each of which is a
separate temporal product market.
7.
Note that customers may benefit from transmission access available for use by
distant generators, even if the local areas has numerous generators that are not able to
exercise market power because of the competition between them. If the transmission
access allows distant generators with lower costs to compete with local higher-cost
generators, customers will benefit from switching to the distant generators and the
efficient outcome may involve the exit of some or all of the local generators. See FERC:
REPLY COMMENTS OF THE NORTH CAROLINA UTILITIES COMMISSION 33 (Dec. 2002),
available at http://www.ncuc.commerce.state.nc.us/electric/rcomments.pdf.
8.
Russell Pittman, Structural Separation to Create Competition? The Case of
Freight Railways, REV. NETWORK ECON., Sept. 2005, at 181, available at
http://www.rnejournal.com/articles/pittman_rne_sept05.pdf (discusses modern evidence of
09HILKE FORMATTED
158
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
[1:1
The primary sources of my disagreements with the article
are two-fold and center around its pessimistic assessment of the
status and prospects for competition in the U.S. electric power
industry.9 First, there are conditions other than the number of
market participants that are critically important in addressing
market power concerns. Second, there are potentially important
approaches to increasing the number of market participants that
are neglected in the article. Both of these avenues of competitive
restructuring have been successfully pursued by the Federal
Energy Regulatory Commission (FERC) and several states. The
net result of considering these two sources of disagreement is
both a more optimistic picture of the state of regulatory reform
efforts in U.S. electric power markets and a more positive view of
FERC’s efforts and those of some states. I am concerned that
Professor Carstensen’s article confounds the inherent
complexities of regulatory reform in electricity markets with the
delays and backtracking in the U.S. reform process instigated at
least in part by strong parochial interests opposed to competitive
electric power markets.10
II. HISTORICAL IRONY
Before detailing my disagreement with Professor
Carstensen, I will elaborate on a historical irony about U.S.
regulatory reform efforts in the electric power industry that
partially explains, as Professor Joskow stated it, “The Difficult
Transition to Competitive Electricity Markets in the U.S.”11
When considering the drawbacks of cost-based rate
regulation or nationalization as policies to address market power
concerns about natural monopolies, economists often focus on the
lack of incentives to minimize costs and to innovate. Some of
these efficiency concerns follow from principal-agent problems
associated with cost-based price regulation or nationalization,
and some of them follow from the related public choice literature.
Faced with a choice between forms of intervention, many
the importance of economies of vertical integration in the rail industry).
9.
I do not discuss here specific disagreements with Professor Carstensen about
historical events and technical conditions that I expressed to him separately.
10.
There are many public interest grounds for concerns about transitioning to
competitive electric power markets such as loss of vertical efficiencies, existing market
power, and lack of demand response, but these concerns are quite different than the
interests opposed to increased competition because it diminishes market power and
provides access to lower cost power to areas that have not had access in the past.
11.
PAUL L. JOSKOW, THE DIFFICULT TRANSITION TO COMPETITIVE ELECTRICITY
MARKETS IN THE U.S. (AEI-Brookings Joint Center for Regulatory Studies July 2003),
available at http://www.econ.cam.ac.uk/electricity/publications/wp/ep28.pdf.
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
159
economists would likely favor cost-based rate regulation over
nationalization, because the efficiency and innovation incentive
problems of nationalized firms can be even more severe than
those facing private regulated utility firms.12 Residual claimants
are more dispersed for nationalized firms than for private
regulated firms,13 and nationalized firms operating as
government agencies are sometimes required to operate in ways
that add to their costs.14 Many economists would, therefore, view
the U.S. as lucky to have avoided the electric utility
nationalization that became the predominant structure in most
of the rest of the world. Ironically, in the transition to
competition, nationalization has been a blessing in disguise,
because it appears to have made the reform process more
politically manageable. In particular, when the government owns
the monopoly, vertical unbundling does not entail private
stranded cost concerns and horizontal disaggregation does not
entail forcing private firms to sell assets that they want to keep.15
In short, the U.S. has a good historical excuse, in addition to
federalism, for its difficulties in bringing competition to
electricity markets. At the same time, whatever the historical
excuses, the U.S. could risk a loss of comparative advantage
(because electricity is a ubiquitous factor of production in traded
goods) if it cannot match successful transitions to competition in
foreign electric power markets that reduce foreign electric power
costs relative to U.S. electric power costs.
12.
The issue of internal efficiency incentives is at the core of the economic
rationales for privatization. See, e.g., JOHN VICKERS & GEORGE YARROW, PRIVATIZATION :
AN ECONOMIC ANALYSIS (MIT Press 1988); PRESIDENT’S COMMISSION ON PRIVATIZATION ,
PRIVATIZATION : TOWARD MORE EFFECTIVE GOVERNMENT (Mar. 1988), available at
http://www.sciencedirect.com/science.
13.
When principals are dispersed, organizing effective monitoring of agents
becomes a public goods problem to the extent that the rewards of monitoring efforts by a
principal (lower costs resulting in high profits) are shared with other principals even if
they do not make a comparable monitoring effort.
14.
A compelling description of X-inefficiency issues and over-staffing issues in the
case of the corporatization of the New Zealand electricity system is B ARRY SPICER ET AL .,
THE POWER TO MANAGE: RESTRUCTURING THE NEW ZEALAND ELECTRICITY D EPARTMENT
AS A STATE-O WNED ENTERPRISE - THE ELECTRICORP EXPERIENCE (Oxford University Press
1991).
15.
It is difficult to envision how the U.K. electricity reforms could have been
initiated in anything like the actual time frame if the institutional complexities of the
U.S. situation had been present. Effectively, the government was willing to absorb the
stranded costs in the form of lower offer prices for its generation assets caused by the
prospect of increasing competition.
09HILKE FORMATTED
160
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
III.
[1:1
COMMENTARY
A. Conditions Other than the Number of Market Participants
Are Critically Important in Addressing Market Power
Concerns
My primary sources of disagreement with Professor
Carstensen remains, however, that electric power markets have
some important and fairly unique properties. These properties
include the impracticability of storing electric power, the physics
of electricity transmission, and persistent economies of scale in
transmission and distribution. Many of the conditions that
prevent exercise of market power in other markets apply to
electric power markets as well. I have termed this the consumer
self-defense perspective on market power in electricity markets.16
Just as in other markets, market power is less likely to be a
problem in electric power markets if (1) there are several
suppliers, (2) consumers have accurate and timely price
information, (3) consumers can quickly and easily switch between
suppliers, (4) consumers—practicably—can “make” instead of buy
the product, (5) inventories are available to supplement current
product when prices are high, or (6) long-term and short-term
supply agreements are available to buyers. When these elements
of a consumer self-defense perspective are considered more
generally, each is a means of making residual demand facing a
supplier more elastic. Each is also an element in antitrust
analysis of horizontal mergers because of potential impacts on
the elasticity of the residual demand facing each supplier. The
price elasticity of residual demand facing each supplier is one of
the primary conditions which determines the profitability of
unilateral anticompetitive price increases (attempts by
individual generators to exercise market power).17
My concern is that Carstensen’s article overemphasizes the
number of suppliers and overlooks other approaches that have an
effect on the residual demand elasticity facing suppliers. I am
somewhat more optimistic about competitive reforms in U.S.
electric power markets because I hinge less of my evaluation of
market power remedies on market share concentration. As in
antitrust enforcement, economists tend to view concentration as
an important indicator, but not the end of the story. The counter
argument is that during peak periods, even firms with small
16.
John C. Hilke, A Consumer Self-Defense Perspective on Electricity Markets, 33
LOY. U. CHI. L.J. 805 (2002).
17.
Another contributor is the level and the shape of a supplier’s cost curves.
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
161
market shares can have market power in electric power markets
with highly inelastic demand and entry impediments, but one
does not have to go all of the way to atomistic competition to
address this concern.
Accordingly, I think it is worthwhile to take a look at the
other items in the list of consumer self-defense conditions and
relate them to recent developments and prospective
developments in U.S. electric power markets.
Consumers have accurate and timely price information:
Nearly any analysis of market power problems in electricity
markets starts with the assessment that average cost retail
pricing results in a market demand that is extremely inelastic.18
This is a condition ideal for the exercise of substantial amounts of
market power—potentially by suppliers controlling a modest
19
proportion of total generation capacity in an area. Many of the
worst potential market power problems in electricity markets
could be substantially relieved by greater price sensitivity of
demand. Even modest reductions in consumption are likely to
produce major price effects during peak demand periods when
generators in the steep section of the supply curve set the market
clearing price.20 One of the most optimistic changes of the past
couple of years is that several states with retail competition are
waking up to the significance of real-time pricing for combating
market power problems in electricity market.21 An important
footnote here is that many of the benefits from real-time pricing
can be gained by exposing large customers to these rates, even if
22
most residential customers continue with averaged rates. The
18.
Stoft explains, “Perhaps the most dramatic structural problem of power markets
is the almost complete lack of demand response to fluctuations in the wholesale price. It is
conceptually dramatic because it sometimes prevents the intersection of the market’s
supply and demand curves, a flaw so fundamental it is not addressed on any economic
texts.” STOFT, supra note 1, at 78.
19.
Capacity in an area that is only available at prices far above the existing market
clearing price does not constrain a small, but significant price increase by the owner of the
generator that sets the market clearing price in that time period.
20.
For example, the Electric Power Research Institute (EPRI) estimated that a
2.3% reduction in consumption during peak days can reduce spot prices by twenty-four
percent. ROBERT LAURITA, ISO N EW ENGLAND 2003 DEMAND RESPONSE PROGRAMS, 6
(May 2003), available at: www.uinet.com/pdfs/ISODemandResponseProgram.pdf; see also
FED. TRADE COMM’N, COMPETITION AND CONSUMER PROTECTION PERSPECTIVES ON
ELECTRIC POWER R EGULATORY R EFORM: FOCUS ON R ETAIL C OMPETITION, ch. III (Sept.
2001), available at http://www.ftc.gov/opa/2001/10/elecretailcomprprt2001.htm.
21.
New Jersey and Maryland are examples. See, e.g., EDWARD J. BLOUSTEIN ,
ASSESSMENT OF CUSTOMER R ESPONSE TO R EAL TIME PRICING 7 (Sept. 6, 2005), available
at http://policy.rutgers.edu.
22.
See, e.g., GEN. A CCOUNTING OFFICE, R EPORT TO THE C HAIRMAN, COMMITTEE ON
GOVERNMENTAL AFFAIRS, U.S. SENATE, ELECTRICITY MARKETS: CONSUMERS COULD
BENEFIT FROM D EMAND PROGRAMS, BUT C HALLENGES R EMAIN (Aug. 2004); C ONSUMER
09HILKE FORMATTED
162
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
[1:1
benefits of real-time pricing can include lower system costs due to
flatter load profiles and more efficient investment incentives for
consumers23 as well as reductions in the market power of
suppliers.24
Consumers can quickly and easily switch between suppliers:
Having multiple suppliers in a market does little to protect
established consumers from market power if switching is costly
or time consuming. If switching is easy and customers have more
attractive offers from other suppliers, efforts to impose
anticompetitive price increases will cause consumers to switch to
other suppliers, thus making the price increase unprofitable. At
the retail level, states are entirely responsible for regulating
switching. States have backed away from some requirements
that produced high switching costs.25 Texas reports that a large
proportion of customers now understand that switching suppliers
26
is relatively quick and easy. At the wholesale level, the costs of
switching include the costs of obtaining transmission service for
a different source of supply. FERC’s open access and RTO
policies have been developed to avoid transmission
discrimination as an element in wholesale switching costs.27
ENERGY COUNCIL OF AM., POSITIONING THE CONSUMER FOR THE FUTURE: A ROADMAP TO
OPTIMAL ELECTRIC POWER SYSTEM (Apr. 2003), available at http://www.cecarf.org/
Publications/MiscPub/RestExecSummary.pdf.
23.
STOFT, supra note 1, § 1.2–1.5.
24.
Id. § 4.1–4.5. A caution here is that if suppliers are exercising market power in
the form of temporal price discrimination, average pricing would represent the single
monopoly price while time-sensitive pricing would represent third degree price
discrimination. The welfare effects of third degree price discrimination are ambiguous.
25.
For example, New Jersey had a “wet signature” requirement whereby a retail
supplier had to confirm a customer’s intention to switch by obtaining a signed agreement
from each customer. Retail competitors found that few customers returned written
agreements and this increased the marketing costs of retail competitors and contributed
to low initial switching rates in New Jersey. Other states found that internet
confirmations or voice recordings were sufficient proof of a customer’s intention to switch
between suppliers. New Jersey subsequently withdrew the wet signature requirement.
See FED. TRADE C OMM’N, COMPETITION AND CONSUMER PROTECTION PERSPECTIVES ON
ELECTRIC POWER REGULATORY REFORM: FOCUS ON RETAIL COMPETITION , ch. V (Sept.
2001), available at http://www.ftc.gov/be/v000009.htm.
26.
A Public Utility Commission (PUC) of Texas survey found that sixty-six percent
of customers viewed the process of switching retail electric power suppliers as easy. Press
Release, Tex. Elec. Choice, Texans Want Competition - In Football, Barbeque and Even
Electricity (Nov. 10, 2003), available at www.powertochoose.org/media/press.asp?aid
=149f&pageid=1 (last visited Feb. 25, 2006).
27.
Recovery of Stranded Costs by Public Utilities and Transmitting Utilities, 61
Fed. Reg. 21,540 (Apr. 24, 1996) (codified at 18 C.F.R. pts. 35 & 385) [hereinafter FERC
Order 888]; Open Access Same-Time Information System (Formerly Real-Time
Information Networks) and Standards of Conduct, 61 Fed. Reg. 21,737 (May 10, 1996)
(codified at 18 C.F.R. pt. 37) [hereinafter FERC Order 889]; Regional Transmission
Organizations, 65 Fed. Reg. 809 (Jan. 6, 2000) (codified at 18 C.F.R. pt. 35) [hereinafter
FERC Order 2000].
AN
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
163
Consumers can practicably “make”—instead of buy—the
product: On-site generation is the most direct option for
consumers to make electric power. Consumers with on-site
generation capability can help the system in several ways. First,
because these customers have a “make” option, they are
particularly price sensitive and can make it substantially more
difficult for generators to profitably impose anticompetitive price
increases. Second, the capacity of on-site generators can provide
generation reserves that increase system reliability. Third, by
flattening system load, the on-site generators allow the system to
increase the proportion of low cost base-load generation capacity
and decrease the proportion of high cost peak-load generation
capacity. FERC has fostered efforts to develop connection
standards for on-site generators as well as for new commercial
generators. States are responsible for other important
components of practicable on-site generation. One is the rates
that on-site generators charge the local distribution utility for the
power they transfer to the grid. Another is the rates that utilities
or other suppliers charge to on-site generators for backup
power.28
Inventories are available to supplement current product when
prices are high: Electric energy storage technologies (other than
the long-established pumped storage approach) are just
developing (e.g. flywheels). The efficient development and
diffusion of these technologies, just as those of on-site generation,
is likely to depend on accurate and timely retail prices for electric
power that is transmitted over the grid. This is largely an area of
state regulation.
Long-term and short-term supply agreements are available to
buyers: One constraint on short-term pricing is the price terms
available under long-term supply agreements. Prices in long29
term supply agreements are constrained by the costs of entry.
Hence, the availability of long-term supply arrangements
associated with entry may help constrain short-term supply
prices. An important task for FERC is to encourage availability
of long-term transmission service agreements to match long-term
generation investments. If a new generator lasts thirty years, but
28.
A useful discussion of on-site generation issues is CONSUMER ENERGY COUNCIL
AM., DISTRIBUTED ENERGY: TOWARDS A 21ST CENTURY INFRASTRUCTURE (Jul. 2001),
available at http://www.nrri.ohio-state.edu/dspace/bitstream/2068/569/1/DE+Final+Repor
+7-01.pdf.
29.
That is, a long-term supply contract with a new generator is a potential
substitute for a long-term supply contract with an existing generator and anticompetitive
price increases by existing generators may prompt entry by new generators or vertical
integration by customers (on-site generation, for example).
OF
09HILKE FORMATTED
164
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
[1:1
transmission service agreements are limited to five years, the
entrant is exposed to the risk that after five years it will not have
sufficient transmission access to supply potential customers at
competitive prices.
B. Other Ways of Increasing the Number of Market Participants
My second source of disagreement is the focus of the paper
(Section II) on horizontal divestiture as the primary means to
increase the number of suppliers and, therefore, to prevent
exercise of market power through unilateral action or
coordinated interaction.
The proposition that horizontal divestiture of generation by
distribution utilities to multiple parties is one method to change
the market structure from a monopoly to one with several
suppliers is self evident. Indeed, states including New York,
California, Texas, and Maine have pursued this policy to various
degrees.30 My point of disagreement is that horizontal divestiture
is not the only way to increase the number of suppliers if there
are too few to facilitate effective competition in some time
periods. In fact, some of FERC’s best work in the past decade has
been directed at increasing the number of suppliers available to
wholesale and retail customers by expanding the geographic
scope of U.S. electricity markets. Reducing transmission rate
pancaking and unbundling generation from transmission to
discourage transmission discrimination (that narrows geographic
markets) have both been actively pursued by FERC.31 When
geographic markets are expanded, more suppliers are able to
32
reach each customer, and supplies are less lumpy because
30.
Massachusetts also ordered divestiture of generation, but allowed a single buyer
to acquire all of a utility’s generation assets. This still allowed competition between
generators because the franchise areas were relatively small and well interconnected.
Divestiture has primarily been used by these states as a method to exactly determine the
level of stranded costs, rather than leaving the assessment to an administrative process.
31.
FERC Order 888, supra note 27; FERC Order 889, supra note 27; FERC Order
2000, supra note 27.
32.
Under the methodology for delineating markets in the DOJ/FTC Horizontal
Merger Guidelines, each time period is a separate product market (because demand and
supply in one period of time are largely independent of demand and supply in other time
periods) and the size of the geographic market is likely to differ in different time periods
as transmission congestion conditions change. In this context, the focus of analysis is on
suppliers that could undermine a small, but significant (and non-transitory or recurring)
price increase imposed by a profit-maximizing hypothetical monopolist of supply in the
proposed market. Hence, not all capacity with transmission access is equally pertinent.
Capacity only available at prices far above the existing market clearing price is not
relevant in that time period and should not be considered in market concentration
calculations. However, this capacity may be pertinent in other time periods when the
market clearing price is higher. See FED . TRADE COMM’N (2000), supra note 1, ch. VI; see
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
165
generators can sell a portion of their output to several other
potential wholesale customers. Similarly, FERC has worked to
make entry less risky by standardizing requirements for
connecting to the transmission grid. Another interesting
approach to increase the number of suppliers is to allow retail
customers to “sell” demand reductions back into the wholesale
market. This has the potential of “killing two birds with one
stone”: such a program would directly increase the price
sensitivity of demand facing incumbent generators and
potentially vastly increase the number of suppliers since
consumers that conserve during demand peaks would become
new suppliers during these periods of time.33
IV. CONCLUSION
In summary, my view is that Professor Carstensen’s article
takes too narrow a view of the basics for competitive markets.
Therefore, the paper misses some of the important good news
from the regulatory reform front in U.S. electricity markets.
When the paper asks: “are we there yet?”, I would not answer
“yes”. I would, however, answer “yes” for the organized markets
of the Northeast and Texas, because we are making discernible
progress. At the top of my “to do” list in these areas are: more
progress on price responsive demand, longer-term financial
transmission rights, backstop transmission investment programs
in the RTOs, and backstop transmission siting authority for
FERC.
Let me conclude with three cautions. First, the good old days
were not so good, and would not translate well into current
conditions. For example, a great deal of independent generation
investment was made in anticipation of low natural gas prices.
However, many of these suppliers are facing difficult financial
conditions as a result of higher natural gas prices. In the absence
of regulatory reform, much of this capacity might well have been
part of the rate base, just as many of the substantial nuclear
plant cost overruns became part of the rate base in the 1970s.
Under the regulatory reforms, consumers have not generally
borne this investment risk as they once did.
Second, going back to traditional regulation would likely
involve its own substantial transition costs. For example,
relaxing the prohibitions on transmission discrimination may
also the addendum.
33.
For example, in the ISO New England Day Ahead Market, customers submit
bids into the Day Ahead Market and are paid for reductions in projected load at the bid
price or the day ahead clearing price. See LAURITA, supra note 20.
09HILKE FORMATTED
166
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
[1:1
force cooperatives and municipal utilities to become more selfsufficient in generation, even if lower priced generation is
otherwise ready and willing to serve them. As a result, this could
be a costly proposition for customers of these organizations.34
Finally, the rest of the world is not standing still and has
already learned from our costly mistakes in California.35 To date,
electricity regulatory reforms in other countries such as the
United Kingdom, Australia, New Zealand, and Chile, have been
quite successful in driving down costs. 36 Also, the European
Union is closing in on a consensus model for electricity market
reforms. To this end, it would seem odd that the country that
successfully championed competitive markets through the cold
war cannot get itself organized well enough to match the success
of competition in this industry among its traditional allies.
V. ADDENDUM
A. Several Suppliers: Why Low Aggregate Concentration Is
Likely to Be Necessary If There Are to Be Two or More
Suppliers Competing to Set the Market Clearing Price in
Each Time Period
Several aspects of the operation of electricity markets lead to
the conclusion that several suppliers are likely to be necessary in
an area for temporal electricity markets to have two or more
suppliers competing to set the market clearing price. These
elements suggest that several suppliers are necessary to avoid
exercise of market power, but this is not the same thing as saying
that atomistic competition is necessary. One potentially
important way to assure that several firms affect the supply
situation is to bring retail demand response into the wholesale
market—this effectively converts many customers into potential
suppliers of “negawatts”—reduced consumption that can
substitute for generation.
Due to retail prices that average the procurement costs of
34.
JOHN C. HILKE, O BSERVATIONS ON THE HISTORY OF TRANSMISSION OPEN A CCESS
R EVISITING THE ISSUE OF EFFICIENCY INCENTIVES FOR R EGIONAL TRANSMISSION
ORGANIZATIONS
(American
Antitrust
Institute
2005),
available
at
http://www.antitrustinstitute.org/recent2/366.pdf.
35.
Primarily these include retail rates that do not adjust for changes in fuel costs,
severe restrictions on hedging activity by retail suppliers, and market rules that make it
easier for suppliers to exercise market power. See John C. Hilke & Michael Wise, Who
Turned Out the Lights? Competition and California’s Power Crisis, ANTITRUST, Summer
2001, at 76.
36.
A good insider’s review of the effects of exposure to competition on utility
operations is SPICER, supra note 14.
AND
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
167
electric power over extended periods of time, consumption of
electric power is more volatile than it would be if retail pricing
tracked wholesale prices over time. Accordingly, retail demand
for electricity is highly inelastic (at least in the short run)
because retail prices do not reflect temporal fluctuations in the
costs of producing electric power.
Wholesale demand at each point in time is entirely derived
from retail demand at that point in time because electric power
cannot be practicably stored in large quantities with existing
technology (no inventories).
Due to the high volatility of consumption and generation
technology that generally makes it less costly to use the full
capacity of one generator before utilizing another generator,
electric power suppliers in a given area use a variety of
technologies and fuel sources with different cost structures.
Depending on the technology used and the fuel used, some
generation units operate nearly all of the time they are
physically capable of doing so (base load) while others operate
only intermittently over the course a year.
In organized electric power markets, such as those within
regional transmission organizations (RTOs), spot markets
organized by the RTO have a central dispatch policy that pays
generators on the basis of the market clearing price and
dispatches all generators with bids equal to or below the market
clearing price in each specific time period, taking into account
transmission costs as well as generators’ bids. The RTOs also
organize markets for ancillary services necessary to maintain the
quality and reliability of the system.
Due to the volatility of consumption and the wide variety of
generators’ marginal costs, the range of bids in electric power
markets is often large. Traditionally, generators are divided into
base-load, mid-merit, and peaking generators. Base-load
generators, such as nuclear plants, have low marginal costs and
cannot be rapidly brought on line. At the other end of the
spectrum, peaking generators have high marginal costs and can
be readily brought on line or taken off line for brief periods of
time.
Because electric power cannot be practicably stored, demand
and supply must be nearly perfectly matched at every point in
time in order to maintain system reliability. As a result, each
time period is effectively a separate product market, with
potential differences in the size of the associated geographic
market based on transmission congestion conditions.
In this situation, not all generators provide a constraint on
09HILKE FORMATTED
168
3/31/2006 9:25:39 AM
ENVIRONMENTAL & ENERGY LAW & POLICY J.
[1:1
efforts by the generator at the margin to raise prices above the
competitive level. For example, if a five percent price increase
above the competitive market clearing price is used in assessing
which generators are in the market, the generators able to
profitably undermine this price increase may be a small fraction
of generators in the area (or with access to the area). That is,
many generators in the area that may be in the market at some
periods of time are not in the market at other periods of time
when the market clearing prices are much lower. Hence, even if
two firms in the area are sufficient to ensure competitive prices
during the period of lowest demand, several generators are likely
to be required in the area to have effective competition at all time
periods. For example, if there are three segments in the supply
curve, six generators would be required in order to have two
generators competing to establish the market clearing price in
each segment. Generally, aggregate concentration in electricity
markets refers to all capacity or output shares at peak demand
when all available generators are dispatched. This contrasts with
the capacity or output shares that are applicable to a market
structure analysis of a specific segment of the supply curve that
will be pertinent when the market clearing price is within that
segment.
B. Vertical Separation: Supply from Distant Generators
Requires Transmission Access
Both transmission and distribution involve substantial
economies
of
scale
making
overbuilds
economically
impracticable.37 State and local governments control siting and
have not generally allowed overbuilds. Under these conditions,
the only way that a distant generator can supply a customer is by
gaining access to the transmission and distribution systems. The
only alternative is on-site generation, but this may be
impracticable for some customers given existing technology and
existing distribution infrastructure for delivery of fuels for onsite generation.
However, a transmission monopolist whose franchise area
includes the prospective customer of the distant generation may
have an incentive and the ability to block or discriminate in
supplying transmission access for such a proposed transaction. In
one scenario, the incumbent transmission owner has incentives
37.
Illustrative figures developed by the Oak Ridge National Laboratory show that
a 765 kV transmission line costs at least thirty percent less than a 500 kV line and at
least eighty-five percent less than a 138 kV line, on a cost per MW-mile basis. FERC
TRANSMISSION TASK FORCE, STAFF R EPORT 215 (1989).
09HILKE FORMATTED
2006]
3/31/2006 9:25:39 AM
COMMENTARY ON CARSTENSEN’S ARTICLE
169
to discriminate against distant generators offering lower prices
because the distant generator reduces the profitability of the
transmission owner’s generation investments either by idling
them or making it more difficult for the transmission owner to
evade rate regulation. In this scenario, transmission and retail
rates are regulated on the basis of costs, plus a rate of return on
investments, and these rates are below prices that would be
charged in the absence of regulation. By buying power from its
unregulated generation affiliate at prices above the competitive
level, the utility is able to increase its profits at its unregulated
generation affiliate and pass these “costs” on through increased
regulated rates. This evades the rate regulation by inflating the
costs of acquiring electric power above the competitive level. The
distant generator in this scenario is a threat to the transmission
owner because its offer may expose the fact that the procurement
costs of the regulated utility are higher than they should be. If
the distant generator displaces or forces a reduction in the prices
paid to the unregulated generation assets of the transmission
owner, the transmission owner’s profits may fall.
In another scenario, regulatory evasion is not required to
motivate the transmission discrimination. In this scenario, the
incumbent transmission owner is already charging the monopoly
price for transmission service and owns higher-cost generation
capacity that may no longer be dispatched if the distant
generator obtains transmission access. If so, the transmission
owner may find it profitable to deny access to the transmission
system or price it at a prohibitive level. It will do so as long as
the gain from protecting the transmission owners’ generation
assets from competition is more than enough to offset the
potential profits from offering transmission services to the
distant generator. This same type of raising rivals’ costs
calculation was involved in the FTC’s concerns about the
proposed merger between PacifiCorp (a utility with extensive
generation assets) and Peabody Coal Company (the sole fuel
source for some large generators operated by rivals of PacifiCorp
that set the market clearing price during some periods of time).38
Utilities’ efforts in areas with relatively high generation costs to
identify and to recover stranded costs were based on the similar
concern that some generation units required in an era of local
generation self-sufficiency would no longer be economically viable
if distant, lower cost generators have transmission access so that
they can compete against the local higher-cost generators.
38.
FED. TRADE COMM’N (2000), supra note 1, ch. 6.